Brand loyalty, once linear, is now fluid, thanks to the rise of LLMs, and heightened expectations of personalisation. But in an era of high churn, razor-thin 2% conversion rates, economic headwinds, and fierce competition, how well are brands really keeping up? Avya Chaudhary mapped what’s actually working. It’s not what you'd expect.

Traditionally, to drive loyalty, brands relied on exclusive offers and perks tied to tiered status programmes (e.g., silver, gold, platinum users based on customers spending power), redeemable points (allocated per dollar spent), subscription-based offers (e.g., free delivery, exclusive deals), and shared values & ideology (e.g., sustainability, inclusivity). 

However, the market dynamics have changed dramatically since the onset of LLM models. ChatGPT and Gemini have started to bulldoze the old customer journey by becoming de facto research assistants and changing how customers move through discovery, consideration, and engagement stages. 

The previous touchpoints used to be a winding road of blogs, reviews, and maybe a friend’s opinion. Now, a new parent asks ChatGPT for the safest convertible car seat, and gets a clear, customised list with prices and links, before even opening Amazon. That’s how fast things are moving. And in such a crowded market, speed and relevance matter more than ever. 72% of people say fast service earns their loyalty. Two-thirds say they’ll stick with brands if they offer more relevant, personalised experiences. 

Brands like Starbucks and Sephora are already putting this into practice by using AI to deliver one-to-one product recommendations at scale. "Both these brands have applied it in genuinely customer-centric ways," says Natasha Wallace, chief solutions officer for strategy & planning at Jellyfish. "By suggesting relevant products through AI, they’ve boosted purchase frequency and encouraged adoption of new items."

It’s reverberant that brands will need more AI-mediated brand loyalty programmes to meet the heightened personalisation demands, and match customer expectations in real-time. 

Brands are betting on AI to make loyalty personal again 

Suzanna Chaplin, founder and CEO of esbconnect, was ahead of the curve in using AI to build customer loyalty, and she hasn’t looked back. Her perspective is refreshingly clear: even with all the noise around machine learning and LLMs, the pillars of customer loyalty haven’t really changed, just the delivery mechanisms. A superior product, joyful user experiences, reliable service, and alignment in values still win hearts. What’s evolving is how we deliver them. 

"AI could be a personal stylist and 3D image of you, so you can try on clothes virtually and get recommendations and no longer need to browse. You simply come on, say what event you are shopping for, and suddenly you are in a virtual changing room with a bunch of outfits."  

This is only a glimpse of what’s possible. More and more brands are now building intelligent ecosystems by stitching together proprietary data from their customer data platform (CDP) or CRM tools with LLM models. Data from POS, purchases, browsing, social media, and support tickets is cleaned and standardised to create 360-degree customer profiles and understand features like transaction recency, product affinity, and sentiment shifts. 

These are fed into ML models that drive timely product recommendations, trigger automated service responses, and generate personalised journeys across email, apps, and websites. The technology is already here. The brands seeing results are the ones who’ve figured out how to use it continuously, not just once.

But as this infrastructure matures, the customer isn't the only audience anymore. AI agents like ChatGPT Operators are beginning to act as intermediaries in the decision-making process. They can sense intent, scan for optimal options, and even autonomously act on behalf of the end customer to make a purchase or stop a subscription. 

When the buyer is a bot and bot has opinions 

Such agents are in a nascent stage and have not yet penetrated the market to the extent where it becomes significantly critical for your revenue. But given the speed at which AI is exploding, the wise move isn’t to wait for the battleground to form, but to lay the groundwork now.

"I believe we’ll see two-system brands, agentic and human," says Wallace. Brands will need to foster loyalty with agents through functional, accessible information, and with people through relevant and entertaining content. "A brand should not think about omni-channel but design a touchpoint ecosystem to serve this split audience, with a confident understanding of where to consolidate energy and impact for people vs agents." 

Sam Panzer, director of industry strategy at Talon.One, extends that logic into the near future. He notes that by 2026, 50% of transactions will involve agents, initially in discovery and recommendation, but increasingly across the full buyer journey. "An AI agent can know a customer so well that it can streamline the entirety of anticipation of needs, discovery and transaction into a single endeavour. It can know a customer so well that they no longer need to shop; it shops for them."

Already, companies like Klarna are designing content and product descriptions specifically optimised for AI assistants: fact-based, structured, machine-readable. In parallel, they are doubling down on human-side engagement via influencer partnerships and experiential loyalty programmes. As Natasha puts it, "Building a brand is even more important than ever, and finding avenues for people to spend time with a brand, like social media, activations and real-world experiences, loyalty programmes, these all become so important to foster connection."

The risks of over-AI’d experiences 

The road to AI-powered personalisation is increasingly complex. Brands must now navigate strict data privacy expectations, avoid emotionally intrusive experiences, and prepare for the growing influence of agentic AI systems capable of making decisions on behalf of consumers. Sam Fenton-Elstone, CEO and co-founder of Anything is Possible, points to last year’s Spotify Wrapped as an example of AI misfiring at scale. "You could sense that the Spotify Wrapped experience had been overengineered with AI, becoming too cute and feeling fake. This was in contrast to the previous years, when Spotify's already impressive tech had struck a better balance." 

That visibility is part of the problem. Consumers are increasingly wary of what brands know and how they know it. With LLMs frequently hallucinating outputs, lacking explainability, and operating on fuzzy security standards, brands risk creating systems that feel invasive or opaque. Say your AI infers that a user is facing financial difficulty based on behavioural patterns, then serves payday loan ads. It might be "accurate," but violates data protection laws like GDPR and undermines trust at scale.

"Brands also need to foster trust with their customers," Fenton-Elstone adds. "The more data they can work with, the more specific a service can be; however, that data needs to be secure, and consumers must be able to trust it. The question marks around many LLMs’ core security settings make AI problematic when it comes to highly sensitive, personal, or financial data."

New customer journeys, new signals, and loyalty-aware KPIs

Of course, the pre-AI era  KPIs of measuring brand loyalty are still relevant: repeat purchase rate, customer lifetime value, net promoter score, redemption rate, churn rate, etc. However, with evolving loyalty programmes, these KPIs too would need to adapt to the agentic ecosystem. 

Going forward, companies might see these traditional KPIs in tandem with zero-click conversions, personalisation/recommendation engagement rate, and omnichannel consistency score. Brand loyalty would mean earning trust from both humans and their AI agents. In the agentic ecosystem, newer indicators such as agentic abandonment rate, trust alignment score, and AI decision efficiency will provide better insight into how AI agents respond to brand experiences.

Panzer points out that great loyalty strategies will become layered as AI explodes. "That can come in the form of gamified experiences that boost interaction, rewards like free shipping or bonus gifts for buying directly, or creating VIP experiences that money can’t buy." Winning brands will be designing seamless, ethical ecosystems that serve emotion, logic, and data, making them the preferred choice across both.